1870.] DOUGLAS — ON OBSERVATIONS OF THE SUN. 141 



a sliOAver towards the sun by tlieir attrition produce a sheet of 

 liaht resemblinu' the corona? Moreover, as the meteors whieh 

 fall upon our earth are composed almost entirely of iron, we may 

 suppose those reaching the sun to contain that metal as a predo- 

 minant element. Although the spectroscopic observations of the 

 corona differ — Professor Young having detected several bright 

 lines, and Professor Harkness only one — by both the presence of 

 iron is rendered highly probable. 



POLARISCOPIC OBSERVATIONS. 



Prof. Pickering entirely disagrees with the observers of the 

 Indian eclipse as to the polariscopic condition of the coronal 

 light. lie says: — "The form of polariscope used was that 

 adopted by Arago in his experiments on sky polarization. It 

 consists of a tube about twenty inches long and two inches in 

 diameter, one end of which is closed by a double-image prism of 

 Iceland spar, and the other by a plate of quartz. Looking 

 through the former, we see two images of the latter, which, when 

 the light is polarized, assume complementary tints. If, now, the 

 corona was polarized in planes passing through the centre of the 

 sun, (as is generally admitted,) when viewed through the polari- 

 scope, in one image the upper and lower parts should have 

 appeared blue, and those on the right and left yellow, while in the 

 second image these colours would be reversed, — the yellow being 

 alone below, and the blue on the sides. In reality the two 

 images were precisely alike, and both pure white ; but one was on 

 a blue, and the other on a yellow back-ground. From this we 

 infer that the corona was unpolarized, or, at least, that the 

 polarization was too slight to be perceptible." Prof. Pickering 

 adds, that "although this does not prove that it shines by its own 

 light, since polarization is produced only by specula, and not by 

 diffused reflection, yet these observations, and those by the 

 spectroscope, seem to render it probable. This view is also 

 strengthened by the fact, that as the most distant portions are but 

 about 100 parts the distance of the earth, they receive about 

 10,000 times as much heat per square foot. The coloured back- 

 ground mentioned above shows that the sky, close to the corona, 

 is strongly polarized; and, since the tint is uniform on all sides of 

 the sun, the plane of polarization is independent of the position of 

 the latter — that is, the same on the sides that it is above and 

 below it. The most probable explanation of this most unexpected 



